Welcome to the World of Advanced Probability!
In the Standard Level (SL) course, you learned how to count outcomes and handle basic probability distributions. Now, we are stepping into the Higher Level (HL) territory. This chapter covers two major areas: Bayes' Theorem (which helps us "reverse" conditional probabilities) and Continuous Random Variables (where we use calculus to find probabilities for measurements like time, weight, or height).
Don't worry if these terms sound intimidating! We will break them down into simple steps using logic and a bit of integration.
1. Bayes' Theorem: Updating Our Beliefs
At its heart, Bayes' Theorem is a way to calculate conditional probability "backwards." Usually, we know the probability of an effect given a cause. Bayes' Theorem lets us find the probability of a cause given an effect.
The Formula
For two events \(A\) and \(B\), the formula is:
\(P(A|B) = \frac{P(B|A)P(A)}{P(B)}\)
According to the HL syllabus, you may need to handle up to three events (e.g., three different machines making parts). In this case, we use the Law of Total Probability to find the denominator \(P(B)\):
\(P(A_1|B) = \frac{P(B|A_1)P(A_1)}{P(B|A_1)P(A_1) + P(B|A_2)P(A_2) + P(B|A_3)P(A_3)}\)
The "Tree Diagram" Hack
If the formula looks scary, use a tree diagram! It is often the safest way to solve these problems:
- Draw the branches for the first set of events (the "causes," e.g., Machine A, B, and C).
- Draw the second set of branches (the "outcomes," e.g., Defective or Not Defective).
- To find \(P(A|Defective)\), calculate: (The branch you want) / (Sum of all branches that lead to Defective).
Quick Tip: Always make sure the probabilities of your initial "cause" branches add up to 1!
Key Takeaway: Bayes' Theorem is just a fraction: (The specific path we are interested in) divided by (The total probability of the outcome occurring across all paths).
2. Variance of Discrete Random Variables
In the SL section, you learned about the Expected Value \(E(X)\), which is the mean. In HL, we also calculate the Variance, which measures how "spread out" the results are.
The Formula
\(Var(X) = E(X^2) - [E(X)]^2\)
To calculate this step-by-step:
- Find \(E(X)\) by multiplying each value \(x\) by its probability \(P(X=x)\) and adding them up: \(E(X) = \sum x P(X=x)\).
- Find \(E(X^2)\) by squaring each value \(x\), multiplying by its probability, and adding them up: \(E(X^2) = \sum x^2 P(X=x)\).
- Subtract the square of your first answer from your second answer.
Standard Deviation: Remember that \(\sigma = \sqrt{Var(X)}\).
3. Continuous Random Variables (CRV)
Unlike discrete variables (like rolling a die, where you can only get 1, 2, 3...), continuous variables can take any value in a range (like height: \(170.532...\) cm). We use a Probability Density Function (PDF), denoted as \(f(x)\), to describe them.
Rules for a Valid PDF
For a function to be a PDF, it must satisfy two rules:
- Non-negative: \(f(x) \ge 0\) for all \(x\). (You can't have negative probability!)
- Total area is 1: The integral over the entire range must equal 1: \(\int_{-\infty}^{\infty} f(x) dx = 1\).
Calculating Probability
In a continuous distribution, the probability of an exact value is zero: \(P(X = 5) = 0\). Instead, we find the probability of being in a range by finding the area under the curve:
\(P(a \le X \le b) = \int_a^b f(x) dx\)
Did you know? Because the probability of an exact point is zero, \(P(X < 5)\) is exactly the same as \(P(X \le 5)\) for continuous variables.
4. Mean, Median, and Mode for CRVs
Just like discrete data, we can find the "center" of a continuous distribution, but we use calculus instead of sums.
The Mean (Expected Value)
\(E(X) = \int_{-\infty}^{\infty} x \cdot f(x) dx\)
Think of this as "weighting" each value of \(x\) by its density and adding them all up.
The Median
The median \(m\) is the value where the area to the left is 0.5:
\(\int_{-\infty}^{m} f(x) dx = 0.5\)
The Mode
The mode is simply the value of \(x\) where \(f(x)\) is at its maximum. To find this, look at the graph or find where \(f'(x) = 0\).
Variance and Standard Deviation
The formula is identical in structure to the discrete version:
\(Var(X) = E(X^2) - [E(X)]^2\)
Where \(E(X^2) = \int_{-\infty}^{\infty} x^2 \cdot f(x) dx\).
Key Takeaway: If a question asks for mean or variance of a continuous variable, get ready to integrate! If it asks for the median, you are solving for the upper limit of an integral.
5. Linear Transformations: \(aX + b\)
What happens to our mean and variance if we change our data? For example, if every student gets 5 extra marks (a shift) or if we double everyone's score (a stretch)?
If we define a new variable \(Y = aX + b\):
- Effect on Mean: \(E(aX + b) = aE(X) + b\)
(The mean is affected by both multiplication and addition.) - Effect on Variance: \(Var(aX + b) = a^2 Var(X)\)
(The variance is only affected by the multiplier, and you must square it! Adding a constant \(b\) does not change how spread out the data is.) - Effect on Standard Deviation: \(\sigma_{aX+b} = |a| \sigma_X\).
Common Mistake: Students often forget to square \(a\) when finding the new variance. If you double the data (\(a=2\)), the variance becomes 4 times larger (\(2^2 = 4\))!
Summary Checklist
- Can you draw a tree diagram to solve a Bayes' Theorem problem with 3 events?
- Do you remember that the total area under a PDF \(f(x)\) must equal 1?
- Can you set up the integral for \(E(X)\) by multiplying \(x\) and \(f(x)\)?
- Do you know that adding a constant (\(+b\)) to a variable doesn't change its variance?
- Can you find the median by setting the integral equal to 0.5?
Note: For help with specific distributions like the Binomial or Normal, please refer to the "Discrete random variables and the binomial distribution" and "The normal distribution and standardization" chapters.